Stars form in large clouds of cold, dense molecular gas. In these clouds the majority of stars do not form in isolation, but instead form in clusters. The formation of stars and their hos~ clusters are intrinsically linked, and thus to fully understand how stars form we must also understand the formation and early evolution of stellar clusters. The formation of stars is thought to be governed by the turbulent conditions inside these molecular clouds, and due to this the initial conditions of star formation are likely to be spatially complex and dynamically cool. In this Thesis we use fractal spatial distributions (D = 1.6,2.0,2.6 and 3.0) to mimic the complex initial conditions of star formation to investigate how the dynamical evolution of star clusters is affected by variations in the amount of primordial structure. We also use varying initial virial ratios (Q = 0.3, 0.4 and 0.5) to investigate what affect the initial kinematics have on a clusters dynamical evolution. I present a new method, based on the minimum spanning tree, which is able to determine and quantify the presence of mass segregation. The method is applied to observations of the ONe, ,vhich we find to be complexly mass segregated, with different levels of mass segregation depending on stellar mass. We find, contrary to common belief, that mass segregation can occur through purely dynamical processes on a short timescale (rv the initial cluster crossing time). We also find that the amount of dynamical mass segregation that occurs is dependant on both the initial structure and virial ratio, where cooler and more structured initial conditions tend to lead to more dramatic dynamical evolution. Additionally, we find that the clumpy and cool initial conditions also lead to the dynamical formation of high-mass multiple systems, which in turn can lead to the ejection of high-mass stars and the destruction of the host cluster itself.